15.1.5 Does not fix 450b SM issues with weaker performance than 450d side by side

Someone needs to address this ASAP, we have 5 units side by side at an apt complex and the 450b sm's are at 16 qam compared to the 450d sm's at 64 qam also with 5-9db difference in rssi.

3 Likes

Yeah, I have seen the same thing.  The problem is worse when the 450b connects to an AP in UNII-2 than UNII-3

we see this as well

NetOpsCom, MSImax & Ryan,

Could you please email me "engineering.cgi" files from 450d and 450b units you have compared side by side.
My email id is balaji.grandhi@cambiumnetworks.com.

We recommend upgrading the firmware to 15.2. We have fixed a ethernet lockup issue on 450b SMs in 15.2.

Thanks,

Balaji

We are also seeing issues with much lower (reported) SNR on a 450b high gain(would it really have been so hard to give the two units different names?), compared to a 450d. It may be due to Rx sensitivity as the SNR tends to jump around a lot but the numbers are certainly different.

We are digging into the SNR reporting differences that folks are seeing... 

If you are seeing poor performance (in addition to the differences in RSL or SNR), please open a support ticket.  We need to collect information, perhaps even get some equipment back, or make a visit to the site so we can get to the bottom of what's going on.

I just upgraded 1 450b high gain to the new release 15.2.1 and still see the weak performance.  Even had some guys go by and repoint to make sure again.  Still weak signals.  1.5 miles and can't even get to 64qam.  we are seeing better signal quality from the 450b mid gain.  Anyone seeing anything better from the 450b high gains?

All, as the originator of this post I feel I must update it with new info. Cambium is fully aware that there is something going on, but its not nailed down as to exactly what. At this point everyone is in agreement that the D and the B are about 3db in difference when comparing rssi. This is semi made up for though because the 450B can transmit higher than the 450D. Notice the D will transmit at 22dbm while the B transmits at 27 dbm. Our rtm came to our network after having this issue sent up to Engineering and even England. Prior to the visit, we had all of our sm's connected to only 450m's. We purchased an antenna aligner for horns and while up a tower one day we checked the 450m and found we had almost -8* down tilt  total when we factored for only -4* total. We adjusted and instantly saw better receive signals across the board. We never went back to some of the original sites we tested however to check alignment and it kind of bit us in the ass when RTM showed up. We were able to test at multiple locations up to 4 miles (because that's our 5ghz cut off) and were able to get within 3 db of rssi between every 450d and 450b we tested. Not sure why the first round we had so much issues or why we saw such a huge swing in rssi between the 2. At 3.8 miles we had a tripod 10' off the ground and were receiving -63 on the 450d and -65 on the 450B. HOWEVER, with all that said, at that same location we did see an issue with SNR that i hope cambium still takes very seriously. The 450d by far at every location we tested had better SNR and a more stable SNR, meaning it was not jumping around and both chains (h and v) were basically the same. The 450b seemed to get worse the farther we got out, which also could explain why people are seeing worse results over 5 miles which seems to be the consensus amongst everyone having this issue. By the time we got to 3.8 miles the 450d was still humming along at 64qam high 30%'s on both chains and staying very steady. The 450b though was actually having issues on the V pol but running strong on the h pol. We saw 16qam on V and a strong 64 qam on h. It was roughly 22db on v and 30-32db on h. This was pretty much line of sight so I would conclude that with a little obstruction the issue would probably become even worse and more apparent to most users. We are running 15.1.5 for those interested. Also, attached are 2 sm's side by side that are actually in use and installed at a multi tenant building so they are roughly 15-20 feet apart. Even here at 2.28 miles, we see the 450b jumping modulation a lot. 

450B

450d

3 Likes

Has there been an update to this issue? We are seeing poor results with 450b high gain. Seems to only be effected when using UNII-2. We cannot connect anyone past 4 miles. We swap them over to a vanilla 450 SM with reflector dish and get -76. Have tried upgrading AP and SM to latest 15.2.1 with same poor results. We've cancelled our order for additional high gain SM's until this can be resolved. 

I hope this can be resolved quickly.  We have several AP deployments that we are holding off on because we are seeing very poor results even at 2 mile range in 5.8.

We've found that the 450b high gain performs quite nicely with our 5.8ghz 450M AP's. I think the tighter uplink on the AP is a good match with the 450b's. However, the 450b's on our vanilla 450 AP's have been quite disapointing. As the noise floor increases, less gain on the 450b over the 450'd and reflectors is not progress.

Last week we changed out all 24 SM's on an AP. 4 of these customers we could not use 450B's and had to use normal 450's. The 450's came up no issue at all with OK signal. These 4 customers were the 4 with the worst signal but still within our guidelines. I was hoping to make this AP 100% 450b's but that will not be possible. This was disappointing.

Ben

While exchanging emails with a Cambium engineer on this, we had a light bulb moment when it comes to 450b SM's on UNII2 frequencies. We were not properly setting the external gain setting on our vanilla 450 SM's. Since the 450b SM's have integrated antennas, the external gain setting is fixed. In short, our 450 SM's with reflector dishes on UNII2 AP's are set to TX over the legal limit. Could this be your experience as well?

2 Likes

Any update on this? Is it a 450b hardware issue?

Last week I was able to do some more testing along-side some Cambium engineers.  We found that the 450b high gains needed quite a bit of adjusting to get them dialed in, but performance once we got there was good.  They are not completely sure why they are so touchy, but they are going to continue to figure that out.  We found that just moving the whole SM up/down, even just a few inches would change the signal as much as 8 dbm, sometimes it was a few feet.

3 Likes